Going Nuclear: The Fusion Race Is Heating Up. Will Anyone Cross the Finish Line? (Video)

A team at Lockheed Martin’s Skunk Works revealed late last year that they’re at work on a truck-sized fusion reactor. Amazon’s Jeff Bezos and others have plugged money into General Fusion in Burnaby, British Columbia. Microsoft co-founder Paul Allen and Venrock have invested in Tri-Alpha Energy, a secretive effort near Irvine, Calif. And Peter Thiel’s Mithril and Y Combinator have funded a Redmond, Wash., startup called Helion Energy.

But the billions of government dollars and decades of research invested in replicating the power source of the sun have, to date, largely resulted in missed deadlines, cost overruns and incremental progress.

As the old joke goes: Fusion power is thirty years away — and always will be.

So there’s considerable skepticism in the scientific community that these private upstarts can achieve such ambitious timelines — or whether they can pull off fusion at all.

“It’s probably a better bet than that Nigerian prince that keeps emailing me, but I would not invest my money in it,” said Edward Morse, a professor in UC Berkeley’s nuclear engineering department.

He said that fusion researchers have tried most of these alternative approaches and simply failed to produce the results needed to justify continued investment.

Morse and others in the field believe the more promising paths today remain the massive, government-backed efforts, notably ITER, a so-called tokamak reactor under construction in southern France, which confines plasma fuel in the shape of a giant donut. But the international scientific collaboration is years behind schedule and estimated costs have more than tripled to around $20 billion.

It would be difficult to overstate the promise of successfully commercializing fusion. It is free from the meltdown dangers of fission, potentially far more efficient than renewables like solar and wind — and unlike fossil fuels, it wouldn’t pump out the greenhouse gases that are warming the planet.

Where nuclear fission splits atoms, fusion pushes two nuclei together, releasing energy in the process.

A single gram of a fusion fuel like deuterium oxide is potentially equivalent to 10 tons of coal, according to Helion. That’s enough fuel to power a home for a year, all packed into a vial no bigger than your pinky.

David Kirtley, chief executive of Helion, believes his startup can build a compact prototype reactor that generates more energy than it consumes within the next three years, achieving what’s known as “scientific gain,” by pursuing an approach known as magneto-inertial fusion. It relies on magnetic fields to hold and compress the plasma fuel.

“We think that we’re transitioning this technology away from billion-dollar scale government programs, to these small private fusion efforts where innovation can build on all of those years of R&D and science, to then move forward with small, distributed fusion systems,” he said.

To learn more, watch the video above.

Update: This story has been updated to clarify that Helion hopes to achieve what’s known as “scientific gain” within three years, not “ignition,” which means a self-sustaining fusion reaction.

This article originally appeared on Recode.net.